如何重命名多个文件中的多个列？

Question

I have multiple files which look like this:

trans_ENSG00000047849.txt.traw
trans_ENSG00000047848.txt.traw
trans_ENSG00000047847.txt.traw
...

In them I have around 300 columns, and column names look like this:

NA20826_NA20826 NA20828_NA20828 NA20819_NA20819

I would like that my column names in all files have instead this form:

NA20826 NA20828 NA20819

In other words I would like to remove everything after underscore _ in every column name and in every file.

I should mention that there is a here is a tab space at the beginning of each file.

I tried this:

sed -ri 's/[_].*$//' trans_*.txt.traw

but when I tried to open one of these transformed files in RI got this error:

> e=read.table("trans_ENSG00000135541.txt.traw", header=TRUE)
Error in read.table("trans_ENSG00000135541.txt.traw", header = TRUE) : 
  more columns than column names

Answer 1

I guess you actually want this:

$ echo -e "\tNA20826_NA20826\tNA20828_NA20828\tNA20819_NA20819" | sed -r '1s/_[^\t]*//g'
        NA20826 NA20828 NA20819

_[^\\t]* since it's TAB separated, so starting from _ to before the TAB (or end of line) are things to be deleted.
g flag is to replace all occurances in line.
The first 1 is to limit the replace in first line -- The title line.

Your own s ubstitude command 's/[_].*$//' , is to replace from the first _ to the end of the line, so it will ends up with only one title left.

Answer 2

Sed command you need is:

sed -ri 's/_\S*//g'

This regexp removes part of every word, starting from underline until next space or tab character, no matter how many columns has each line.